This pop-sci piece extols the virtues of intermittent fasting, though the author gives it weight over calorie restriction that it doesn't merit at this time. The evidence is much stronger for the benefits of calorie restriction, as it has been studied more extensively. The results for extended longevity in laboratory animals due to intermittent fasting remain mixed, though it certainly seems to produce health benefits:
One of the most important studies in this area was conducted just last year at Salk's Regulatory Biology Laboratory. In an experiment, biologist Satchidananda Panda and colleagues restricted the feeding of mice to - conveniently enough - an 8-hour period each day. The researchers were attempting to study whether obesity and metabolic diseases like diabetes were the result of high-fat diets, or from the disruption of metabolic cycles.
To that end, Panda gave the mice lots of fat to eat. In fact, 60% of the calories consumed were derived from fat (which was meant to simulate foods like chips and ice-cream). The researchers also created a control group that ate the same thing, but these mice could eat any time they wanted (interestingly, as nocturnal creatures, they ate half their meals at night, while grazing on the remainders during the day). As for the restricted group, their 8-hour window was at night.
One hundred days later, the free-for-all group was a mess. They gained weight, developed high cholesterol, high blood glucose, and experienced liver damage and diminished motor control. But as for the mice who practiced the intermittent fast, they weighed 28% less and showed no signs of adverse health.