CR 174: Case-Control Study of Motorcycle Crashes (1997)


This report presents the findings of the Case-control study of motorcycle crashes. The cases were 222 motorcycle crashes occurring on public roads in the Melbourne metropolitan area from late November 1995 to 30 January 1997 in which the rider or pillion was taken to one of the participating hospitals or died. The controls were 1195 motorcyclist trips which passed the crash site at the same time of day and day of week as the crash occurred.

The study collected three types of information:

  • detailed descriptive information about the crash and the resultant injuries
  • comparison of features of cases and controls, and
  • motorcycle exposure information (gathered as part of collection of control data).

Characteristics of Cases

Of the 222 crashes, 22 involved pillions. Twenty-two riders and three pillions were killed in the crashes, which had the following characteristics:

  • most commonly occurred on Fridays
  • generally highest frequencies from noon to 8 pm
  • almost 20% occurred near the centre of Melbourne
  • 80% in urban areas
  • almost half were on major arterials
  • 65% occurred in 60 km/h zones
  • more than two-thirds on curves
  • equally divided between intersection and non-intersection locations
  • mostly on two-way undivided roads
  • very few local area traffic management devices at crash sites
  • the road was not clean at almost one-quarter of the sites and there was deformed pavement or a sudden change in road surface at many sites
  • about half occurred on two-lane roads
  • poles, kerbs and trees were present at most sites
  • there was no evidence of braking at 85% of sites
  • 9% occurred when it was raining
  • about one-quarter occurred under difficult lighting conditions (glare, dusk or dawn, night-time)
  • sun glare could have reduced visibility at 13% of sites
  • glare from oncoming headlights was a potential problem at 8% of sites

Type of crash

  • one-third were single vehicle crashes
  • two-thirds of all crashes involved impact with an object or vehicle, in half of all crashes this was a moving car
  • single vehicle crashes were more likely than multi-vehicle crashes to involve alcohol, to occur at night and to involve excessive speed
  • 23% of crashes were judged to have involved excessive speed for the conditions
  • the rider was judged to have contributed to about two-thirds of the multi-vehicle crashes, mainly by inappropriate positioning or failure to respond
  • most riders did not consider themselves to be at fault in multi-vehicle crashes to which failure to respond was judged to contribute


  • 167 motorcycles were inspected
  • 15% had travelled less than 5,000 kms
  • more than 20% were judged to be not well cared for (dirt and mud etc)
  • almost 40% were judged to have been in a poor to fair mechanical condition (compression, bearings etc.) prior to the crash
  • clean motorcycles were mostly in good or excellent mechanical condition, whereas most of the motorcycles in poor mechanical condition were dirty
  • about a quarter had under-inflated front or rear tyres
  • a quarter had a worn or loose chain
  • 15% had brakes in a poor condition, typically insufficient pad thickness
  • 19% of rear tyres and 7% of front tyres were badly worn or bald


  • 145 helmets worn by riders and pillions were inspected
  • over 50% were black or "dark"
  • 20% of visors were tinted
  • the average age was four years, with 16% more than 5 years old and so may no longer have been performing optimally
  • more than 80% had obvious signs of damage, mostly scratches but some fractures
  • in 43% the interior padding was visibly worn or compressed

Injuries to Motorcyclists in Non-fatal Crashes

  • the median Injury Severity Score (ISS) was greater for admitted motorcyclists than presentations (10 versus 5)
  • 4% of all injured motorcyclists had severe head injuries
  • 3 of the 5 motorcyclists not wearing helmets sustained head injuries
  • facial injuries were uncommon and not significantly more common among those wearing open face helmets than full face helmets (8% versus 4%)
  • chest injuries were uncommon but relatively severe when they occurred
  • 44% of motorcyclists had upper limb injuries and 57% had lower limb injuries
  • most common injuries overall were fracture of the knee or lower leg (28%) and fracture of the forearm (17%)
  • external injuries (abrasions, contusions or lacerations) occurred to 88% of motorcyclists but were generally not severe
  • there was no clear indication of differences in injury severity for riders and pillions in the same, non-fatal crashes
  • single and multi-vehicle crashes did not differ in their injury severity
  • injury severity and patterns of injury did not vary significantly as a function of speed zone
  • wearing appropriate clothing did not significantly decrease the likelihood or severity of external injuries

Case-Control Comparisons

The data from 205 cases for which controls could be recruited and 1195 controls were compared. Where odds ratios are cited in this section, they are statistically significant.

Rider factors

The factors which were associated with significantly increased crash risk after adjustment for potential confounding factors were:

  • age under 25 (compared with age 35 or over)
  • never married
  • unlicensed
  • experienced off-road rider before gaining on-road licence
  • having fewer years of on-road riding experience (after adjustment for BAC>.00)
  • ride less than 3 days per week - this may be an artefact of the study design
  • having completed a beginner course compared with an advanced course
  • BAC>.05 (odds ratio of 38) - 13% of crashed riders for whom BAC was known had BAC>.05 compared with less than 1% of control riders
  • BAC>.00 (odds ratio of 5)
  • consumed alcohol in the previous 12 hours (odds ratio of 2)
  • not wearing a helmet (2% of crashed riders and 1% of controls)

Other results included:

  • 6% of crashed riders and 3% of control riders had used illicit drugs (mainly marijuana) in the previous 12 hours but the unadjusted and adjusted odds ratios were not statistically significant
  • 11% of crashed riders and 8% of control riders had taken prescription drugs in the previous 12 hours but none of the odds ratios were statistically significant
  • 2% of crashed riders and 3% of control riders had taken nonprescription drugs in the previous 12 hours but the numbers were too small to analyse
  • after adjustment for BAC, there was no significant increase in risk associated with wearing an open face helmet compared to a full face helmet
  • no significant increase in risk associated with wearing a helmet 5 to 10 years old or over 10 years old (compared with one less than 5 years old)
  • no significant increase in risk associated with wearing a helmet that did not belong to the rider after adjustment for age and BAC
  • none of the odds ratios associated with wearing protective gear were significantly different from one. However, these analyses were based on self-report data for cases and observation for controls and so may have been affected by a social desirability bias for cases.

Pillion factors

The presence of pillions could possibly contribute to either crash causation (e.g. by behaving inappropriately or simply by producing a higher centre of gravity) or increased crash severity (because they are another person who may be killed or injured).

  • pillions were present in 10% of crashes and 7% of controls
  • significant increase in crash risk associated with pillion carriage
  • 70% of pillions in crashes were female, 57% of control pillions

Motorcycle factors

The factors which were found to significantly increase crash risk after adjustment for potential confounding factors were:

  • riding a motorcycle with engine capacity of 750 cc and above compared to one of 260 cc or below (adjusted for licence status)
  • the rider not being the owner of the motorcycle

Other results included:

  • 5% of crashed motorcycles and 1% of control motorcycles were unregistered
  • most motorcycles were manufactured in 1990 or later and so age of the motorcycle varied little between cases and controls
  • two-stroke race replicas comprised 24% of the crashed 250 cc motorcycles compared with 9% of the control 250 cc motorcycles. The increased risk associated with riding these bikes was not significant after adjusting for the effect of alcohol but the adjusted odds ratio was still relatively high (2.7)
  • headlights were on for most of the crashed and control motorcycles (both pre- and post-1992) - the odds ratios associated with pre-1992 motorcycles having headlights off were not statistically significant

Trip factors

  • a significant increase in risk was associated with nonwork-related trips compared with work-related trips
  • no significant increase in risk was associated with the rider being unfamiliar with the road
  • the percentage of riders who estimated that they were travelling at above the speed limit was less at higher speed limits
  • riders with BAC>.000 were more than twice as likely to state that they were travelling over the speed limit

The risk factors for which the contribution to crashes was greatest were:

  • rider aged under 25
  • BAC>.05
  • BAC>.00
  • unlicensed or unregistered or not ridden by the owner
  • nonwork-related riding

It should be noted, however, that the analyses might have overestimated the risk associated with unlicensed or unregistered or not ridden by the owner if these riders were less likely to stop at the control sites than other riders.

Exposure Information

Exposure estimates were calculated based on observations of 1121 motorcycles during 325 hours of sampling at or near the crash sites. The overall proportion of the traffic comprised by motorcycles was very low, about 0.5%.

The highest average number of motorcycles per hour was found on primary arterials (4.05), with the smallest number being found on collector roads (1.23). The proportion of the traffic which motorcycles comprise appears to be similar across road types.

Average motorcycles per hour was greatest during weekday and weekend days and the proportion of traffic that motorcycles comprise was highest on weekend days. Both the average number of motorcycles per hour and the proportion of traffic which motorcycles comprise were lower at night than during the day.

The mean number of motorcycles per hour accounted for 79% of the variance in the number of crashes (per time period or per road type). The regression equation describing the relationship in this particular study is:

Number of crashes=6.50 x mean number of motorcycles per hour

The median distance travelled per week was between 201 and 300 kilometres. Riders holding probationary and full licences rode further per week, on average, than holders of learner permits. Engine capacity per se showed little effect on distance ridden.

The comparisons of recent Australian motorcycle exposure studies showed general agreement although some differences occurred because of the different population base or the time of year that the survey was conducted. In general, the previous studies agreed with the current study’s findings that non-novices rode further per week than novices and that greater distances were travelled by larger capacity motorcycles (greater than 500 cc or greater than 750 cc) than smaller capacity motorcycles.

The data presented here do not clearly indicate any changes in the exposure patterns of riders from 1988 to 1996. While the ABS Survey of Motor Vehicle Use showed that the number of registered motorcycles in Australia decreased from 1985 to 1988, the numbers have remained almost constant since then.

Riding Styles and Strategies

While there were some differences identified, in general, the riding styles and strategies adopted were similar across rider age groups, experience, licence status and training history. The observed differences are summarised below. The greater likelihood that younger riders, many of whom were not fully licensed, had completed at least one training course complicated the interpretation of the observed differences somewhat.

Summary of observed differences in riding styles and strategies:

observational skills frequency of looking behind over one shoulder decreased with age group, more common with training
approaching intersection inexperienced riders more likely to decrease speed, trained riders more likely to change position to improve visibility
following distance longer gap for inexperienced riders, shorter gap for 25 to 34 year old riders
response to tailgating learner and probationary riders and riders with training less likely to speed up
using the horn more by experienced riders
dealing with emergency situations more near misses usually experienced per month by experienced riders (who ride more), youngest age group report most usual number of near misses per month, inexperienced and trained riders more likely to have practised emergency braking and/or counter-steering in the last six months, riders with probationary licences were the most confident about performing sudden swerves in emergency situations
usefulness of training riders aged 35 and over most likely to use cornering skills learnt in training "always"

Download Complete Document: Mcycle_Crash_1 [PDFPDF: 844 KB]

Type: Research and Analysis Report
Sub Type: Consultant Report
Author(s): N Haworth, R Smith, I Brunen & N Pronk
Topics: Crash data, Motorcycle
Publication Date: 01/12/97


Last Updated: 6 May, 2013