Scott Patterson had his faith in the stock market severely shaken over the past year, so he decided to look around for something more solid in which to invest his retirement savings. The 43-year-old owner of a gutter-cleaning business settled last month on a three-bedroom house in Athens, Ga., which over the years has attracted a steady stream of renters from the nearby University of Georgia campus. Patterson put down 20% of the $127,000 sale price, or $25,400. After debt service, taxes, insurance and an allowance for repairs he'll clear $3,500 this year. That's a 12% return on the capital he invested (including closing costs). Depreciation deductions can shelter a good part of this from taxes.

"This just seems like a better bet right now than adding money to the IRA or SEP," he says of his retirement account alternatives.

Until the housing bubble filled investors' heads with the notion that real estate is a way to get rich quickly, the nation was populated by millions of investors, like Patterson, who valued properties based on the rent they could earn. With real estate prices down 50% or more in some areas, the landlord game is back in vogue. And like Patterson, many potential players are carefully comparing it with the returns they could reasonably expect by sticking their money in other places.

Blue-chip stocks? A $100,000 stake invested in the S&P 500 ten years ago would have shriveled to $63,000, including reinvested dividends. Alas, even after a dismal decade stocks are far from cheap in relation to their near-term earnings prospects.

In many metropolitan markets, by contrast, returns from rental properties are at the high end of historic ranges. One standard measure is the capitalization rate, which is calculated by dividing the operating income from a property (annual rent minus upkeep, insurance and management costs) by the sale price. This is comparable with the dividend yield on a stock. For real estate almost anywhere in the U.S., cap rates are better than the current 2.4% yield on the stock market.