Crop-livestock integration is thought to increase system autonomy and resilience by transitioning system components into a circular economy. Like most agroecological practices, crop-livestock integration is considered time-and labor-intensive. We hypothesized that work characteristics (organization, duration, and arduousness) and subjective factors related to work with animals shaped the implementation of crop-livestock integration practices in mixed crop-livestock systems and vice versa. The Qualification and Evaluation of Work method (QuaeWork) was coupled with qualitative interviews. This framework was implemented on 14 farms selected from a typology of mixed crop-livestock systems in Guadeloupe (French West Indies), where the high cost of labor raises tension between work organization and agroecological practices. Three patterns of crop-livestock integration shaped work organization: family farms with strong crop-livestock integration (Pattern 1, 6 farms), farms with moderate crop-livestock integration (Pattern 2, 5 farms), and intensive productivity-oriented farms (Pattern 3, 3 farms). These patterns have different farm characteristics and work organizations. Pattern 2 spent more time using crop by-products as animal feed (on average 253 h/year) than Pattern 1 (on average 222 h/year), whereas Pattern 3 did not use crop-products to feed animals. Farms in Pattern 3 spent more time using crop excreta to fertilize crops (on average 29.8 h/year) than Pattern 2 (on average 19.6 h/year) and Pattern 1 (on average 13.4 h/year). In Patterns 1 and 2, the low availability of family labor, and lack of the presence of skilled employees, may hold back whole-farm crop-livestock integration development and increase in animal units, whereas, in Pattern 3, management of the nutritional value of crop by-products and the cost related to employing additional persons would do so. The subjective link to animals can be more relational (Pattern 1), practical (Pattern 2), or economic (Patterns 2 and 3) regarding farmers' objectives. Adapted mechanization, direct collection of feed and deposition of feces by animals, targeted supportive policies and market governance, and organizational innovation beyond the farm level are discussed as levers that can be operated depending on the pattern.