After pushing upward in late March and early April, the stock market has been losing steam lately.

Among technical analysts -- the folks who try to forecast the market by measuring things such as stock supply and demand and investor attitudes -- worries are spreading that a significant pullback could be in the offing. Some even see signs that a bear market -- a 20 percent decline -- could come either later this year or early next.

Like any other form of market analysis, technical analysis is hardly an exact science, and things can change. But the theory behind technical analysis is fairly simple.

When a bear market is ending and a bull market is beginning, investors are shellshocked and nervous. Stock prices have fallen and people are just getting the courage to jump back in. People slowly buy the beaten-down stocks and prices rise. Optimism rises. Stocks become more expensive, and, eventually, people begin to think about cashing in. Demand weakens, desire to sell begins to outweigh desire to buy, and a bear market slowly begins.

Technical analysts differ from fundamental analysts, who look inside companies at their profits, sales, cash flow and projected product demand. Technical analysts look outside the companies at the market for their stocks, at the people who are buying and selling, and at the amount of stock that is available.

What many of these analysts are seeing is that buying pressure is beginning to wane, and selling pressure is mounting.

"I am not ready to call it a bear market yet, but it does look as though the market wants to take a break," says Louise Yamada, long the head technical analyst at Citigroup who today has her own New York research firm, Louise Yamada Technical Research Advisors. She expects stocks to pull back this spring, although she suspects that they will recover and won't enter a real bear market until later this year or next year.

Ms. Yamada says she was feeling bullish until quite recently. In the past few days, she says, she has noticed the number of advancing stocks slowly being overtaken by the number of declining stocks. In addition, the ratio of companies hitting new highs to those hitting new lows peaked last summer and has been fading. Stocks increasingly are being bought on declining prices rather than rising prices. All these indicators suggest that demand for stocks is flagging.

The bull market "is starting to show a little bit of frayed edges," Ms. Yamada says.

Some technical analysts have been surprised at the durability of the current bull market, after wrongly expecting that it would be over by now. But in some ways, the market has shown signs of age for some time.

Bull markets typically are at their strongest when they are young, and this one was no exception. After the long 2000-2002 bear market ended in October 2002, the Dow Jones Industrial Average jumped 25 percent in 2003. But then, it was up only 3 percent in 2004, and last year it fell slightly, less than 1 percent. So far this year, it's up 3.9 percent.

Although various indexes hit multiyear highs in March and April, the gains have been limited.

A pullback this spring would reflect the usual seasonal pattern, says Phil Roth, chief technical market analyst at New York brokerage firm Miller Tabak. Stocks typically start the year strong, pull back around now as investors take profits, recover some ground in the summer, and then suffer another decline in the September-November period, he says.

He expects the market to follow that pattern again this year, with indexes moving in earnest into a true bear market in the latter part of the year. Stock demand isn't looking weak enough yet for a bear market to progress very far this spring and summer, he says. "It is looking like the bigger damage comes at the end of the year, the September-November period," he says, adding that the November election could cause some additional volatility.

But the economy is proving stronger than many investors expected, which is creating inflation fears and forcing interest rates higher, threatening to pinch profits, he notes. "This is what the end of a bull market looks like: strong economy, rising interest rates," Mr. Roth says.

Ms. Yamada says it is normal for bull markets to end after around four years, and this one will be four years old in October. But in some cases, she notes, bull markets have extended into a fifth year, as happened from 1982 through 1987. She thinks the current bull market could extend into 2007.

A bear market certainly isn't inevitable in that period. Stocks made it from late 1990 through early 2000 without a 20 percent decline in the Dow industrials. But that was an unusual period, and many believe it isn't likely to be repeated.

Paul Desmond, president of Lowry's Reports, a research firm in North Palm Beach, Fla., recently finished a study showing that stocks behave in a surprisingly consistent way when they are topping out. The gains become quite narrow, meaning that index advances are driven by fewer and fewer stocks, as more stocks begin to decline.

In the past, when the Dow industrials were hitting a top and beginning a bear market, an average of just 6 percent of the Dow's 30 component stocks were making highs along with the overall Dow, Mr. Desmond found. An average of 22 percent of the component stocks already were down 20 percent or more from their highs.

Today, the Dow stocks seem to be moving in that direction, Mr. Desmond says. Last summer, when the Dow industrials hit a four-month high, 25 percent of its component stocks made similar highs at the same time. But when the Dow hit its high in March of this year, only about 10 percent of its components were at highs. That isn't quite down to the 6 percent average Mr. Desmond found at the start of past bear markets, which means that the bull market may not be over, in his view.

"We see ourselves as in the final stages of a bull market," he says. "It is conceivable that the Dow hasn't made its final high yet, but the number of stocks going along with that new high will be smaller and smaller all the time."

Last week, although the overall industrial average rose slightly, only 12 of its component stocks showed any gain at all for the week, and 18 were down.

Since many investors own specific stocks rather than broad indexes, Mr. Desmond adds, that means people should be prepared for the possibility that, even if the broad indexes move to multiyear highs in the next few months, their individual stock holdings may not follow suit.