mention 'non-informative' censoring nor that such condition is necessary for the analyses to be bearing on the hazards of interest. Some statements are over-simplified without caution. No variance is given for the K-M estimators. There is no discussion of the specific alternatives over time against which the power of logrank type tests is great, nor of the crossing hazards situation which they may not be able to detect; no discussion either of sample size calculations. Less importantly, no referral to martingale residual plots, repeated events, frailty models, accelerated failure time models, what to do with missing data... In summary, I agree with the author that 'this text provides a straightforward and easy-to-follow introduction to the main concepts', but because of omissions regarding some crucial concepts I hesitate to recommend it as a stand-alone introduction to survival analysis. Opportunities have been missed to provide the novice with a deeper intuitive grasp of the special challenges and simple solutions provided by classical survival analysis, and to guide him or her into the vast literature on this topic. However, for those wishing to obtain a quick understanding of how to read computer output, building on what they know from different regression models, this book will provide valuable help. Others may find in this book useful details and exercises that help to practice independently what they have learned elsewhere.