A survey enrolling 814 schoolchildren was conducted immediately after a peak epidemic in the 1988–1989 influenza season and then a case-control study for influenza-like illness (ILI) was carried Out using information concerning illness onset and usual lifestyle. Based on the analysis of the correlations among symptoms and actions taken due to symptoms, cases were defined as those with fever 38°C and subsequent absenteeism and medical consultation during peak epidemic; within this, mild-ILI (MILI) was defined as fever ≥38°C and ≤39°C, and severe-ILI (SILI) as fever ≥39°C. Controls were defined as those with no symptoms (NS) during that period. Adjusted odds ratios (OR) were obtained through an unconditional logistic regression model between the MILI (80) or SILI (48) and NS (196) groups from among respondents (803). For MILI, increased risks were observed for easily-inflamed tonsils (OR = 3.0, 95% CI: 1.7–5.4); and larger family size (1.9, 1.1–3.4); with decreased risks for higher school grades (0.4, 0.2–0.9); frequent intake of vegetables or fruits other than green/yellow vegetables (0.5, 0.3–1.0); and larger room space per capita (0.4, 0.2–0.9). For SILI, there were increased risks for easily-inflamed tonsils (3.8, 1.8–8.1) and history of doctor- diagnosed asthma (2.9, 1.2–6.7); and decreased risks for higher grades 10, 2, 0.1–0.6), frequent intake of milk or dairy products (0.3, 0.1–0.6) and vaccination 10.3, 0.1–0.8).