The handicap principle has been one of the most important developments in sexual selection theory. Numerous theoretical papers have considered whether extreme male ornaments could have arisen through female choice by being handicaps to the male. These models have been defined as belonging to one of three categories: (i) Zahavi's (J. theor. Biol. 53, 205 (1975)); (ii) the revealing; or (iii) the condition-dependent handicap. Here I discuss whether the division is still helpful for empirical purposes, or whether handicaps are not best considered as being condition dependent. I argue that in fact the revealing and condition-dependent handicaps are indistinguishable empirically, and that the `Zahavi' handicap models arose due to a misunderstanding of what Zahavi originally proposed, which was in fact a condition-dependent handicap.