Following the screening consultation, baseline measures will be collected by the researcher. All children will have baseline anthropometric data recorded (height, weight, BMI, and body fat using bioelectrical impedance). They will then undergo fitness testing. This will consist of tests of grip strength and isokinetic strength testing to assess quadriceps and hamstrings strength in the non-dominant leg. In addition a VO2peak test will be performed using a treadmill protocol. All subjects or their parents/guardians will complete the Haemo-QoL questionnaire  and the Habitual Activity Estimation Scale (HAES) . The HAES questionnaire will be used to explain effects of exercise, should they be apparent, and to determine whether completing a structured 12 week exercise program increases levels of habitual activity.
A randomisation schedule will be created with computer-generated random numbers by a person not otherwise involved in the recruitment process. The allocation will be in randomly permuted blocks of 4, 6 and 8. Allocations will be placed in sequentially numbered, sealed, opaque envelopes to ensure concealment from the researchers recruiting subjects, and from the subjects themselves, prior to randomisation. Potential subjects (or the guardians of subjects) will be invited to participate in the study and the methods of the study will be explained to them verbally and in writing. If they agree to participate baseline measures will be obtained before the subject is randomised to a group. Subjects will be considered to have entered the trial at the time the envelope containing their allocation is opened.
Participants in the control group will receive usual medical care by their treating haematologist and associated health practitioners. Their management will include factor therapy (prophylaxis or "on demand"), as determined by their treating haematologist. All parents will have been educated about the disease and been taught how to recognise a bleeding episode. For the purposes of this study when defining a bleeding episode we have accepted the definition widely used by other researchers that is, a bleed is "an episode of bleeding requiring treatment with clotting factor". Parents will be encouraged to treat bleeding episodes as soon as recognized and advice about duration of therapy will be given by their treating haematologist.
Participants in the exercise group will receive a supervised, general exercise program in addition to their usual medical care. The exercise program will be conducted for 1 hour, two times per week for 12 weeks. The program will consist of a supervised "circuit training" programme including warm-up and warm-down and an age-appropriate mix of resistance exercises and aerobic stations. Aerobic training will consist of 30 minutes of continuous exercise at 60–70% HRmax, depending on baseline fitness of the participant, as measured by VO2peak test. Resistance training will follow the traditional "set system" and involve 20 minutes of upper and lower limb exercises using isotonic pin-loaded equipment. The training load will be set at the 8–12 repetition maximum (RM). Subjects will perform 3 sets, each to failure, with the 8–12 RM load. Throughout the training period the load will be incremented so that failure occurs at between 8 and 12 repetitions. Ten minutes per session will involve warm-up and warm-down exercises. The exercise sessions will be changed frequently to avoid boredom. Patients receiving prophylactic Factor VIII will administer this prior to the exercise intervention. A qualified paediatric physical therapist will conduct and supervise the exercise program. The therapist will undergo additional training in the aetiology and management of haemophilia and understand the early management of a bleeding episode.
Participants in the control group will be advised not to change their physical activity levels during the course of the trial. In cases where this is unavoidable, a record of the changes will be kept. Several mechanisms will be used to ensure that the trial protocol is consistently applied. Protocol manuals will be developed and the researchers (haematologists, physical therapists, exercise physiologists, other outcome assessors and the trial manager) will be trained to ensure that screening, outcome assessment, allocation and treatment protocols are conducted according to the protocol. An independent researcher will randomly audit assessment and treatment sessions to determine adherence to the protocol. Retraining will be offered to treatment providers and outcome assessors not complying with the protocol.
The primary outcomes will be quality of life, strength and aerobic fitness. The details are as follows:
Quality of Life using the Haemo-QoL questionnaire. This questionnaire has been validated in children with haemophilia 
Quadriceps strength using a Cybex isokinetic dynamometer 
using the Bruce treadmill protocol 
The secondary outcomes are as follows:
Incidence of haemarthroses and intramuscular bleeds – self-reported for 16 weeks (12 week intervention period and 4 weeks following)
Lean tissue mass, body fat measured using bioelectrical impedance on a body composition analyser
Habitual activity using HAES questionnaire 
Other strength measures
Adverse events (other than haemarthroses) associated with exercise or physical activity. These will be elicited by open-ended questioning.
All measurements will be taken immediately after the 12 week intervention period, by an assessor blind to group allocation. (The exception is measurement of bleeding episodes, which will be monitored as described above.) Habitual activity and quality of life will be re-measured in a telephone interview at six months. The number and type of any adverse effects that occur for the duration of the study will also be recorded using open-ended questioning.
In order to improve compliance and reduce loss to follow-up, participants in the intervention and control groups will be telephoned at the beginning of each week. Participants in the both groups will be asked about bleeding episodes in the previous week. In addition, participants in the intervention group will be reminded of their exercise training schedule. All participants will be asked to keep a diary to record all bleeding episodes during and for one month following the intervention period. The final data on number of bleeds will be a composite of the data obtained from diaries and from telephone follow-ups.
Sample size and power analysis
A sample size of 70 subjects (35 subjects per group) will provide an 80% probability of detecting an effect of 0.75 standard deviations on H-QOL scale assuming α of 0.05 and worst-case loss to follow up of 15%.
Data will be analysed by a statistician who is blinded to group. All analyses will be by intention to treat. That is, outcome data will be collected regardless of compliance with the trial protocol, and each patient's data will be analysed in the group to which the subject was allocated. The effect of exercise will be estimated with analysis of covariance using a regression approach. To maximise the precision of estimates, pre-test scores of outcome variables will be entered into the regression models as covariates. The uncertainty associated with estimates of the size of effects of exercise will be quantified with 95% confidence intervals. The integrity of the trial data will be monitored by regular scrutiny of data sheets for omissions and errors. Data will be double entered and the source of any inconsistencies will be explored and resolved.