There is of course no such thing as slipstream fiction. The concept is defined negatively, as that which does not fit into defined groups, so one can no more set out to write Slipstream than to write Miscellaneous.
Nonetheless, there are works that one puts in that category. The late 60s/early 70s seemed a time for books that were not exactly "sf" or "mainstream," in that they satisfied the same urge to speculate and explore new concepts that sf feeds, but avoided the usual sf tropes. Some I particularly enjoyed took up Hal Clement's suggestion to consider something obvious and ask what if it were false, but applied the technique to the social, rather than the scientific, sphere.
Brian Aldiss attempted to annex some of these works under the rubric of "lifestyle sf." His prime example was The Dice Man, by Luke Rhinehart, a fascinating meditation on the possibility of radically changing one's life by making important decisions on a random basis. I am told that the book was taken up in sf circles, with at least one fannish group actually trying some "diceliving." Rhinehart followed it a few years later with The Book of est, a lightly fictionalized treatment of the experience of undergoing that briefly fashionable form of personal change. I enjoyed the book and found it credible, but I speak as one who never actually did est. Given the national propensity for forgetting old fads, one might be tempted to reissue The Book of est as a slipstream novel of a somewhat bizarre-sounding imagined therapy. The denial of bathroom privileges would fascinate some readers, but that may be the only thing about the actual est that is generally remembered.
Philip Roth, with his usual chutzpah, wrote a book entitled The Great American Novel, recounting the history of a third major baseball league that we have all agreed to forget. The book is a large and wondrous stew of exaggerated elements from baseball lore (midgets, Black Sox, etc.) and actual American life (McCarthyism), with a sportswriter narrator named Word Smith and players who bear the names of deities (Gil Gamesh, Hothead Ptah, Frenchy Astarte). Though I loved the book, I must admit it was full of wretched excesses, including bad taste, repetition, exaggeration, and general heavy-handedness, none of which should be too shocking to the seasoned sf reader.
Poet David R. Slavitt had set out to become a successful Showbiz Trash novelist under the name of Henry Sutton. He made brief inroads into the best-seller lists, then decided to switch back to novels he was willing to sign his real name to. Perhaps the best of these was The Outer Mongolian, in which a boy with Down's Syndrome is inadvertently turned into a genius by vitamin overdoses. (My exceedingly permissive Suspension of Scientific Disbelief mechanism winced, but grudgingly allowed the explanation to pass.) The lad then Changes the World, becoming the secret cause of many of the political events of 1968. The alternative explanations are, as such things should be, utter impossibilities that one cannot entirely dismiss, and the book has much wit and pathos.
Ishmael Reed has straight-facedly described his Mumbo-Jumbo as a mystery novel, one that should have won the Edgar Award. There are elements of that, with a detective whose name (Papa LaBas) combines Santeria and French decadence, but it is even more interesting as a Secret History, with much being achieved by undercover agencies, conspiracies, and third-world deities. Reed's Flight to Canada is more like an alternate history; its title refers to the escape of slaves from the Confederacy by airplane, and the book's finest set piece is the live telecast of Lincoln's assassination. But rather than assuming a point of divergence from our consensus reality as the subgenre usually does, Reed derives his approach from the belief of some African religions that all time occurs at once.
The books I have mentioned thus far are the sort of "slipstream" published as general or unadjectived fiction, but offering at least some of what sf readers seek. There is, however, no reason why a book could not be published as Science Fiction, then cross the other way, pandering to the desire for verbal wit and gameplaying, self-reference, and other qualities we think of as mainstream, even post-modern.
Which brings me to John Sladek, whose first two novels, The Reproductive System (also published as Mechasm) and The Muller-Fokker Effect, always seemed like the sort of thing that could have been published as literary fiction. (In fact, if memory serves, the first American hardcover of The Muller-Fokker Effect, like that of Robert Sheckley's Mindswap a few years earlier, was presented as an imaginative work of satire, with no reference to how its images resembled those of the books with rocket ships on the covers. Neither book escaped its author's origins; the first paperback edition of each looked science-fictiony.)
In any event, Sladek's books are marvelous. The Reproductive System told the old story of machines that rebel, but made it new, and funny. The Muller-Fokker Effect dealt with computer consciousness (another contender in the great First Cyberpunk sweepstakes?) with equal imagination and wit. The first features a female protagonist whose IQ is so high that her community put her in a school with all the other "special" children because it couldn't think of anything else to do with her; the second includes a men's-magazine publisher whose staff keeps him from ever having sex, on the assumption that it is his endless, virginally desperate lusts that give the magazine whatever helps it sell better than all the other men's mags. There are a few of the now-laughable sfnal commonplaces of the time (such as brand-name marijuana), but that seems a minor quibble. Both books delight with both conceptual inventiveness and word games, literary reference, and paradox.