Joblessness is highly seasonal. To analyze how households adapt to seasonal joblessness, we introduce a measure of seasonal work interruptions premised on the idea that a seasonal worker will tend to exit employment around the same time each year. We show that an excess share of prime-age US workers experience recurrent separations spaced exactly 12 months apart. These separations coincide with aggregate seasonal downturns and are concentrated in seasonally volatile industries. Examining workers most prone to seasonal work interruptions, we find that these workers incur large earnings losses during the off-season. Lost earnings are (i) driven mainly by repeated separations from the same employer; (ii) not recouped at other firms; (iii) partly offset by unemployment benefits; and (iv) amplified by concurrent drops in partners’ earnings. On net, household income falls by about 80 cents for each $1 lost in own earnings.