Observers reported which of as many as eight computer-generated approaching objects would "hit" them first. Accuracy was above chance probability except when two-object displays contained pictorial relative size information that contradicted relative time-to-contact (TIC) information. Mean d' and response time was greater, but mean efficiency (Barlow, 1978)was smaller with eight objects than with two. Performance was less effective when global expansion contradicted TIC information than when local expansion contradicted TIC. Results suggest that observers can judge relative TIC with as many as eight objects when certain sources of information are consistent with TIC and that observers rely on information other than, or in conjunction with, optical TIC. Also, the sources of visual information that affect performance may vary with set size, and identification (but not detection) judgments may be constrained by limited-capacity processing.Information in the optic array specifies when an object will hit or pass the observation point and when it will hit another designated object (Bootsma & Oudejans, 1993;Lee, 1974; Tresilian, 1990 Tresilian, , 1991. Results suggest that observers can use or are sensitive to such optical information about time-to-contact (TTC; e.g., tau, Lee, 1974)1 in various tasks (see, e.g., Bootsma & Oudejans, 1993;Kaiser & Mowafy, 1993;Schiff& Detwiler, 1979;Schiff& Oldak, 1990;Todd, 1981).For example, Schiff and colleagues (Schiff & Detwiler, 1979;Schiff & Oldak, 1990) measured judgments about TTC with a prediction-motion (PM) task in which a filmed or animated object approached the observation point and then disappeared from view. The observers pressed a button when they thought the object would have reached them had the object continued approaching at the same speed after it disappeared. Estimates ofTTC increased as actual TTC increased, suggesting that judgments were based on optical TTC. Moreover, in another study, when observers judged which of two computersimulated approaching objects would contact the observation point first, they achieved over 90% accuracy when there was only a 150-msec difference in TTC between the objects (Todd, 1981; see also Simpson, 1988).Parts of this study were reported at the 1994 annual meeting of the Association for Research in Vision and Ophthalmology. We thank James R. Tresilian for suggesting the analyses of efficiency and local and global expansion. We are grateful to Norma V. S. Graham for valuable discussions and suggestions regarding data analyses and signal detection analyses. We also thank Robert W. Bell, Lanie A. Dornier, Mary K. Kaiser, David 1. Law, and an anonymous reviewer for helpful comments on earlier drafts, and Jeff Maresh and Robert Todd of Engineering Solutions, Inc., for technical assistance. This research was supported by the Research Enhancement Fund at Texas Tech University. Correspondence should be addressed to P.R. Del.ucia, Texas Tech University, Department of Psychology, Lubbock, TX 79409-2051 (e-mail: djxpd@ttacs.ttu.edu).However...