Observations of horizontal velocities just outside the surf zone on two natural beaches show unexpected time lags between the onshore/offshore currents and the longshore currents at incident wave frequencies. These time lags, typically around 20 s, appear to have been first noticed by Katoh (1981) in data from a beach in Japan. A number of possible hypotheses to explain these time lags are examined. Of these, propagation of secondary waves formed by reflection or diffraction at a point alongshore from the measurement site seems most able to explain the observations. This suggests a significance for reflected/diffracted waves, some distance from breakwaters or headlands, which has not previously been recognized in field studies. However, a number of questions remain that will only be resolved by further field data, particularly from beaches far removed from end effects.