Conventional standard-gamble and time-tradeoff methods may be inappropriate for eliciting preferences for acute, nonfatal health states because both require subjects to evaluate tradeoffs between a morbid health state and death. OBJECTIVES: This study demonstrates the use of conjoint analysis with a discrete-choice experimental design to obtain estimates of healthy-days equivalents (HDEs) for clinically relevant durations and severities for rotavirus gastroenteritis, an acute diarrheal illness common in young children. METHODS: Parents of children ≤5 years of age completed a web-enabled survey instrument that presented a series of trade-off questions to elicit judgments about the relative seriousness of pairs of hypothetical gastrointestinal illnesses.  Each illness was defined by seven attributes:  diarrhea, vomiting, fever, seizure, behavioral changes, dehydration, and illness duration. A random-parameters ordered-probit model was developed to estimate the preference weights for each symptom-severity level. RESULTS: A total of 229 parents completed the survey. Subjects judged seizure to be the worst rotavirus symptom, followed by dehydration, vomiting, fever, and diarrhea.  Parents rated behavioral changes as the least bothersome symptom of rotavirus. Estimated preference weights indicate improvements in each symptom relative to its worst level.  As the duration of the illness increases, so does the decrement in HDEs.  The decreases in time equivalents for 3-day and  7-day illnesses with seizure are 0.7 and 2.3 HDEs, respectively, while the corresponding decreases for illnesses with diarrhea are 0.3 and 1.0 HDE, respectively.  These results imply that seizure is subjectively about 2.3 times more important than nine loose bowel movements per day. CONCLUSIONS: Results indicate that the trade-off data support valid estimates of HDEs for acute conditions ranging  in duration from one day to two weeks without requiring subjects to make  implausible comparisons between acute, self-limiting conditions and mortality risks or longevity reductions.The results can be used for cost-effectiveness analysis in place of conventional time-equivalents.