three methods of analyzing catch-effort data were investigated regarding their relationship to walleye abundance: mean cue, gini coefficient and the relative frequency of zero cue (rfz). walleye recreational fishery data were obtained for escanaba lake, wisconsin (1966 to 1985) and the bay of quinte, ontario (1985 to 1989). a linear regression model was used to analyse the relationship of mean cue, gini coefficient and rfz to walleye abundance in escanaba lake. mean cue was not linearly related to walleye abundance. the gini coefficient was weakly correlated with abundance. rfz correlation with walleye abundance was highly significant. only rfz was capable of accurately predicting walleye abundance levels.