CONTEXT: Evolving concussion diagnosis/management tools and guidelines make Knowledge Transfer and Exchange (KTE) to practitioners challenging. OBJECTIVE: Identify sports concussion knowledge base and practise patterns in two family physician populations; explore current/preferred methods of KTE. DESIGN: A cross-sectional study. SETTING: Family physicians in Alberta, Canada (CAN) and North/South Dakota, USA. PARTICIPANTS: CAN physicians were recruited by mail: 2.5% response rate (80/3154); US physicians through a database: 20% response rate (109/545). INTERVENTION/INSTRUMENT: Online survey. MAIN AND SECONDARY OUTCOME MEASURES: Diagnosis/management strategies for concussions, and current/preferred KTE. RESULTS: Main reported aetiologies: sports/recreation (52.5% CAN); organised sports (76.5% US). Most physicians used clinical examination (93.8% CAN, 88.1% US); far fewer used the Sport Concussion Assessment Tool (SCAT1/SCAT2) and balance testing. More US physicians initially used concussion-grading scales (26.7% vs 8.8% CAN, p=0.002); computerised neurocognitive testing (19.8% vs 1.3% CAN; p<0.001) and Standardised Assessment of Concussion (SAC) (21.8% vs 7.5% CAN; p=0.008). Most prescribed physical rest (83.8% CAN, 75.5% US), while fewer recommended cognitive rest (47.5% CAN, 28.4% US; p=0.008). Return-to-play decisions were based primarily on clinical examination (89.1% US, 73.8% CAN; p=0.007); US physicians relied more on neurocognitive testing (29.7% vs 5.0% CAN; p<0.001) and recognised guidelines (63.4% vs 23.8% CAN; p<0.001). One-third of Canadian physicians received KTE from colleagues, websites and medical school training. Leading KTE preferences included Continuing Medical Education (CME) courses and online CME. CONCLUSIONS: Existing published recommendations regarding diagnosis/management of concussion are not always translated into practise, particularly the recommendation for cognitive rest; predicating enhanced, innovative CME initiatives.